 Good afternoon everyone. I hope I'm come after coffee. Hopefully I can keep up with you because probably you have more energy than I do big thanks to Red Hat for having me over for the talk and Today what I will be talking about is completely different than everyone else Everyone has one slide around adoption. I have a full deck around the adoption so I'm Jamil Mina from BMO financial group and I'm the catalyst behind the container adoptions at BMO financial group We're a 200 year old company. We're young compared to Barclay We have 12 million customers 46,000 employees and a lot and a lot of technical debt at BMO We're geographically located across the world mostly our head Head offices in Toronto. We have offices in Asia in Europe and we have a large bank that we own in Chicagoland Harris Bank So this presentation is about container adoption and about the the Ingredients that go in into container adoptions and I'm gonna list the ingredients quickly then we'll throw out the Presentations we're gonna go through them First one the primary ingredient that I look at is a team dedicated team for for adoption The second one is an effort that is focused on container adoption Partner that you can count on which is an important piece of the puzzle in terms of the adoption Keeping it simple failsafe environment and promote and advertise across the organization secure and operational and Tangible consumption model model and finally measuring and tracking the adoption our journey started back in 2016 where we Did a bake-off on container platform and we made and I took the word out here a bet because as a bank We can't bet we have to be secure with your money. We made a bet on Kubernetes an open shift in 2017 we worked with what I call an innovator and I'm gonna expand Innovator in the next slides and so forth to help promote the capability within the organization and 2018 we were planning to go with two applications to production We have three in production. We have major applications launching in June-July It's open shift is the threat strategic platform for our digital transformation We have 10. I think every week the number keeps increasing I think I'm up to 15 applications in the pipelines that are making their way to production now We have hundreds of developers that are trained on open shift and we haven't fully operational and secure platform What surprised us the most is when we launched our expectation it was to have only two applications It was our practice ground one small application and one large application to serve our customers in the United States for our online banking platform and We experienced a hockey stick of demand and we had more demand that we can supply and we had to React quickly and step up to be able to be able to support the demand for it. How did that happen? First and for we talked about the dedicated team my team is responsible for mentoring and Training people enabling people on continuous delivery Within the organization so we have all binary management. We have ansible for automated deployment We have database automation and we have container deployment even though within the organization They know us as the open shift team even though that's not officially our our name that team itself is responsible for Enabling everybody within the organization enabling the operations Enabling the information security Enabling the engineers enabling the developers and building the DevOps pipeline for continuous delivery delivery within the application and training developers on the pipeline The second piece is focus When we looked at the spectrum of container adoption, especially when we're doing the initial bake off We made the conscious choice that our transformation is about the adoption of container is not about the changing of the behavior of the developer one of the things with some With other platforms you need to change the behavior of the developer so the they either want to focus on changing the Pipeline of the developer changing that architecture and getting them to developer to develop a stateless applications Or you can I get the platform like Kubernetes and open shift, which what I call it cloud with training wheels Where developers do not have to refactor that application completely to make them Stateless what's interesting during that process even though the development team wanted something with training wheels They ended up writing applications that are stateless to take advantage of auto scaling auto recovery and so forth that open shift offers Partnering with an app dev Innovator and here that's that's an important concept because you need a strong partner and I'm using here The model that was described back in the 2000 by Jeffrey Moore Which is crossing the chasm and the innovator is that market group. They are opinion leaders and people look at them when they're selecting a platform and The app the app dev the innovator everybody pretends to be an app dev innovator first of all So how do you discover? Who's the app dev innovator? It's someone that is not asking for a business case someone that's gonna go by gut is someone that is looking For is willing to experiment is willing to fail It's willing to invest their best assets, which are that people into your project to see if it's going to be successful Once you find that app dev innovator you cherish them and you start Building your use cases around them and here. This is a busy slide and this purpose of the slide This was a news that came out in the newsletter in a couple of weeks ago from the app dev the innovator Publishing to the whole organization the value that they're getting from the open ships So they start speaking on your behalf and start Communicating to the rest of the organization the value that they are getting and everybody starts to follow at that point because they're looking at that open that the opinion leader and now your head the the mass market of of the users When engineering the the platform keep it keep it simple the first version doesn't have to be complicated Create an MVP that is functional functional functional reliable usable convenient Pleasurable and meaningful, but it doesn't have to have everything Implementing an open shift platform for cloud platform, especially for a regulated organization like us Requires a lot of things that have to be done to ensure that is secure. It's scalable It's has disaster recovery in it has backup has monitoring have to integrate it into your own processes Do not go to the extent just meet the needs of the innovator and what the innovator wanted from us is a CICD pipeline that they can provision environments quickly deploy their application quickly and see how it scale and Would not a lot of refactoring and we did that for them Then the next step we we invested our own money I belong in the DevOps organization and we put $100,000 for development of an application to run with an open shift So by providing the funding we are not under pressure from the project teams or Strategic business alignment for example business breathing down our neck to say I want this function tomorrow in production It allows us space to experiments to fail to recover to learn And that actually that application never made it to production This application was just a training ground and that's where we use it to showcase and promote once we did this application we actually what we did is We did a video and we filmed all the innovators Across the organization and we played that video for the rest of the organization To show them what we've done on on that sample application We were not worried about spending the money if a video cost five thousand dollars to develop a professionally made video We went ahead and we did that Then the next step what we've done is we did a competition. I call it here a hackathon I'm gonna dig in a little bit into it in the next slide and we ensured that we had a competition where people participated and Did something that is meaningful for the organization and then we gave everybody incentives such as prizes Laptops money travels as a result of Of that of the competition and that was spread as news across the organization So the hackathon like event Considering that we are company that manages for 47,000 people that have 47,000 people We could not make people work 24 hours a day So that does not meet the labor code So we're not allowed to put them in the same room and say go at it and create something that is creative and meaningful So we wanted something that works for the organization works for the employees So the way we structured it is we did two sessions of training one session of training Which was around how to use the platform second session of training is how to create an application that works within the platform all the attendees of the hackathon where It was mandatory for them to attend the sessions Then on the day of the hackathon we gave them Some restriction more restrictions for them to use they had to use our existing pipeline They didn't have a choice to use any pipeline that they want to use so we use bit bucket we use bamboo We use JIRA so they had to use our pipeline Second thing is we didn't tell them to create an application from scratch We told them pick any application from the one that you're working on that you're comfortable with Refactor it and make it work within the OpenShift platform and at the end of the day The the participant presented to a group of senior leaders We had all the CIOs the skeptics and the people and the believers in the room from the CIO level to come And judge the platform and that was quite amazing because we did Developers did stuff that the executive did not expect. I actually didn't expect they Delivered the application that were able to do auto-scaling blue green showed logging into our elk platform straight into our platform and Actually, it's a young two graduates from university that won the prize And one of the CIOs asked them how do we do it today in terms when we go to production How do we do the fallback to the previous version and they looked at them? It's like we don't we don't have the capability and I think at that point that CIO Committed all of his platforms to go in on OpenShift To help and resolve that pain and that's how this major application committed to this platform We rewarded the winners with We wanted to send them to a conference a conference they selected to have some nice laptops instead of going to the conference and then we created professional video and We sent that video across the organization and was played in town halls About the event which now we have more and more people starting to ask us They want to be on the platform like everyone else because they've seen the the result Real work started. So now we had to secure to make it secure and operational So we brought more engineers on to engineer the production platform to make sure that we have good networking we have a good the our capabilities and Being a regulated organization We have two two groups that support our environments one non-prod and one a production Group so I took the production people and told them for the next six months I took actually I got one they gave me one. I didn't take any That came and worked with our team and he was embedded with our team to learn how to operate the cluster in a non-production environment, so all of a sudden I had one person From an ops team that is fully capable at operating the platform in production That was our first confidence boot that boost that now if we go to a production with an application We have someone in ops that understands it. They know how to recover an application. They know how to deploy the application They know how to recover OpenShift. They know how to patch it. They know how to secure it and so forth Information security was the other people we had to work with and we actually created a brand new security standard for OpenShift and that's that's that took a six month to create a new standard then create a backlog to apply all the standards against OpenShift and what's interesting with that standard anyone wants to bring in Container platform capability within the organization was measured against that standard How are you going to secure it? How are you going to recover it? How are you going to patch it and so forth? So they had to live by these standards The next step after that was how do you order OpenShift? We have an elaborate Model financial model within the bank for our investment plan for For app dev we had to give people that are used to managing pets How to plan and order carols? So pets is your server which has a name you treat it well, and you know what it is to now It's a cattle. It's a farm of nodes that you don't know that they exist so we created a model small medium large and If a CIO decides that application needs a medium a medium would have four VMs in production One at main site the other one at the DR site two VMs in In the performance environments and six VMs in non-prod environment It was a standard costing for a medium anyone asking medium. That's all what they're gonna get there is no customization to it and then This and then we label these nodes for them to use them the other thing we did is the fast rock for development where we made that the Provisioning of the open-shift cluster a frictionless So when a development team decides to get in on an open-shift, I have a set of nodes within the non-dev environment I can get them to start working that open-shift On the open-shift platform and developing that application which was faster for them Then going and getting VMs creating an environment using VMs would take about Six to seven weeks at the least With open-shift while their costing is happening. They can start developing within two to three days That attracted the dev team to start using the platform as well because now all of a sudden they can fast-track their projects By three to six six months at times to get going And finally is the allocation model Getting the the the culture transition from pets to cattle now you don't own a pet you own a farm of cattle and you can use that farm of cattle's For your own use to scale your application the next evolution of that might be a flat Nodes where they can scale all their application, but we're not there in terms of the culture yet Last but not least is measuring and tracks So we have Dashboard that we've built was pot fire where it gives us at any one time the number of paws that are running That is visible from a business sense that executives can take a look at it at all time And this is published actually to the executives across the organization They can see how many projects we have how many builds we're doing actually there is an anomaly here because There is a team is building so much that they had to clean They would have had more than 14,000 bills in the last in the last year But we we get them to keep purging their their bills that big line on The top right chart is quite interesting because that big line is one of the project teams had 13 environments when in the VM world When they started using the OpenShift platform, they want to about 68 projects the equivalent of 68 Environments and what's interesting is they're using three VMs to run what they used to run on 50 VMs so we experienced a huge density in there in terms of utilization So still having a minute and 38 seconds if there are any questions Happy to take them. You have a great model for for getting buying across multiple levels of the organization. You obviously had some great Support, I'm curious about how you were able to achieve that level of support from the other teams and leaders that you obviously had it from and what Steps were necessary to maintain such a high level of support as you went through the various phases Yeah, so I'm gonna use a buzzword probably is the power of the community We we did have executive support to be able to invest into the promotion and the advertisement But to be honest with you I had to take my foot off the gas Because the adoption rate based on the word within the community that was flowing around between the various leaders Was catching fire faster than my ability to support them I had to slow down a little bit from the advertisement. So I had the other problem I have more demand than I had supply and actually I had to step up I had to add three more people to my team to be able to support the supply before Getting back into advertising. So it was the community that did it for us based on Following the the model of having the innovator to speak to us instead of us speaking for for the product for the service Thank you very much. Thank you